Generally, when defining tables for storage in existing databases or other types of storage mechanisms, a set of records is defined upon creation of a table, with each record including a finite number of fields. Generally, once a table having an established number of fields in the table and a finite number of fields per record is defined, data storage space sufficient to satisfy the data storage requirements of all fields of all records defined in this table are permanently allocated thereto.
A data storage efficiency issue arises with this existing table definition format. Specifically, certain fields within such table may be quite sparsely populated. Otherwise stated, data blocks for some fields may be filled or may contain data for only a small proportion of the records. Data storage space is nevertheless generally permanently allocated for such data blocks, thereby leading to a substantially inefficient use of data storage space.
An example is considered in which a data table of baseball team records associates a record with each player on a team and which assigns fields to represent characteristics of the players. For example, fields could be assigned for such characteristics as height, weight, RBIs (Runs Batted In), hitting percentage, and the like. One field of interest for the present example would be that of an indication of the number of “no-hitter” games pitched by each player. However, for the majority of players who are not pitchers, this field is inapplicable. Moreover, since data storage space was nevertheless allocated to this field, this inapplicable field would generally incur a substantial waste of storage space for those players who are not pitchers.
Therefore, it is a problem in the art that the inclusion of sparse fields in data tables causes the wastefull allocation of space to data blocks which will generally not store useful data.
It is a further problem in the art that recording data pertinent to only a small number of records in a table generally involves incorporating such field into every record in such table, thereby incurring the waste of storage space associated with sparse fields.